Fiber-lasers are gradually replacing conventional solid-state lasers in several laser applications. Fiber-lasers have advantages over solid-state lasers in ruggedness and optical efficiency. CW fiber-lasers are capable of delivering a very high-powered beam, for example, a beam having a power in excess of 1 kilowatt (kW). Pulsed fiber-lasers can deliver peak-power as high as 10 kW or greater. Fiber-lasers can have a high optical efficiency, for example between about 60% and 90%.
High-power CW fiber-lasers with multimode output are extremely useful in material processing applications, such as cutting of complex 3D shapes found in hydro-formed automotive parts and long-offset welding of complex-shaped parts. High peak-power pulsed fiber-lasers with single mode output can be used for scribing of solar-cell panels. Advantageously, high peak-power enables efficient frequency-conversion into visible and UV wavelength ranges.
In theory at least, the output power of a fiber-laser is limited only by how much optical pumping power can be delivered into an optical gain-fiber for energizing a doped-core of the gain-fiber. In practice, there are limits due, inter alia, to non-linear effects which can broaden the spectrum of pump radiation resulting in reduction of absorption efficiency, and photo-darkening of the fiber material which can lead to reduction of efficiency, excessive heating, and even catastrophic failure. The non-linear effects become increasingly problematical as the gain-fiber is longer. Long gain-fibers are necessary with low brightness diode-laser pump sources currently available.
FIG. 1 schematically illustrates a prior-art fiber-laser arrangement 10. Laser 10 includes a gain-fiber 12 having a doped core (not shown). Pump radiation from a plurality of diode-laser modules 18 is coupled into the cladding of the gain fiber via N-to-1 couplers 20 spliced to the gain fiber. Only two diode-laser modules per coupler are depicted in FIG. 1 for simplicity of illustration. In practice there be as many as 6 diode lasers inputting to a 6-to-1 coupler. Pump radiation is coupled into both ends of the gain-fiber.
A resonant cavity extending through the gain-fiber is formed by fiber Bragg gratings (FBGs) 14 and 16 written in passive fibers 15 spliced to the central fiber of the coupler. FBG 14 is maximally reflecting at a lasing wavelength of the gain-fiber and FBG 16 is partially transmissive at that wavelength to allow laser output. The output may be delivered for use in an application or passed on to one or more stages of amplification. This arrangement would require fiber splices (depicted by a bold “X” in FIG. 1) between the diode-laser modules and the couplers, between the couplers and the gain fiber, and between the couplers and the FBG fibers.
Fiber-splices and FBGs can be a source of instability due to transverse mode-coupling. Other issues include grating walk-off and modal instability. The latter issue arises because the fibre laser community, for the most part, is focussed on single-transverse-mode operation of the fiber-lasers in spite of the fibres themselves being multimode to avoid nonlinear impairments. Any fiber-splice is potentially a source of loss, due to less-than-perfect core-alignment, and potentially a source of mechanical failure. Clearly, the more splices the greater will be the potential for problems resulting from the aforementioned issues.
A method of pumping a gain-fiber which does not require fiber splices is to directly focus radiation from an array of diode-laser emitters into the gain-fiber. A one-dimensional array of diode-laser emitters is typically referred to as a diode-laser bar. The emitters have an emitting aperture about 1 micrometer (μm) high (in what is referred to as the fast-axis of the emitter) and a width from about 10 μm to over 100 μm (in what is referred to as the slow-axis of the emitter). The bars are usually about 1 centimeter (cm) long and between about 1 and 4 millimeters (mm) wide, with the emitters having a length in the width-direction of the bar and emitting apertures aligned in the slow-axis direction. Typical diode-laser bars include about 20 emitters with a fill factor of about 20%. If more radiation is required than can be provided by a one-dimensional array, a two-dimensional array of emitters can be formed by stacking a plurality of diode-laser arrays, one above the other in the fast-axis direction, but the separation between bars in the stacking direction is usually greater than about 1.5 millimeters (mm) to allow for each bar to be mounted on a thermally conductive sub-mount for cooling. This provides an aggregate beam which has a radially asymmetric cross-section, being much longer in the fast-axis direction than in the slow-axis direction.
Two-dimensional arrays of this kind can have as many as twenty diode-laser bars vertically stacked providing a total output of a few kilowatts. Such arrays are typically used for heat-treatment of metals and the like where accurate focusing is not required and radial asymmetry is not a problem. This radial asymmetry, however, makes focusing into a gain-fiber difficult and inefficient at best.
In U.S. Pre-Grant Publication No. 2010/0260210 gain-fiber pumping method is described wherein a plurality of diode-laser bars is used to optically pump a corresponding plurality of external-cavity vertically-emitting optically-pumped semiconductor (OPS) lasers with radiation from the OPS-lasers being used to directly pump a gain-fiber. FIG. 2 schematically illustrates a simplified arrangement 22 for carrying out this method. Here, gain-fiber 12 includes FBGs 14 and 16 forming a resonator as described above. Gain-fiber 12 has a doped core 17 surrounded by an inner cladding 19 which is surrounded by an outer cladding 21.
Optical pump radiation is provided by a pump module 23 including plurality of OPS-lasers 24. Each laser delivers a beam of radiation 25 preferably in a single lateral mode or at least a “low-M2” (for example M2<2) mode. The beams are collimated, and are directed parallel to each other, here, by an arrangement of turning mirrors 27, to a positive lens 28. Radiation from all of the beams is focused by lens 28, as indicated by converging rays 29, into inner cladding 19 of gain-fiber 16, with a small portion, of course, directed into core 17. It is taught that in practice, as many as two-hundred fifty beams having M2<2 may be directed onto lens 28 and focused into a gain-fiber having an inner cladding diameter of about 100 μm and a numerical aperture (NA) of about 0.22. Assuming a relatively modest output power of about 30 W for a single-chip OPS laser, it is possible to couple as much as 7.5 kW of radiation into such a gain-fiber.
This pumping method uses the OPS-lasers essentially as “brightness converters” to convert poor-quality low-brightness diode-laser beams into high quality, high brightness laser beams. It remains to be seen, however, whether the advantage of efficient focusing of the OPS-laser beams is sufficient to offset the less-than-100% efficiency of conversion, the cost of the OPS-lasers, and the cost of a beam-combining arrangement for as many as two-hundred fifty OPS laser-beams. There remains a need to develop an effective method of direct pumping using two-dimensional diode-laser arrays.